A study of the magnetotransport properties of the graphene (II. Fractional Quantum Hall Effect).pdf
-
Upload
dwipesh-majumder -
Category
Documents
-
view
4 -
download
0
Transcript of A study of the magnetotransport properties of the graphene (II. Fractional Quantum Hall Effect).pdf
Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/280099788
Astudyofthemagnetotransportpropertiesofthegraphene(II.FractionalQuantumHallEffect)
RESEARCH·JULY2015
DOI:10.13140/RG.2.1.3737.9684
READS
13
1AUTHOR:
M.A.Hidalgo
UniversityofAlcalá
63PUBLICATIONS634CITATIONS
SEEPROFILE
Allin-textreferencesunderlinedinbluearelinkedtopublicationsonResearchGate,
lettingyouaccessandreadthemimmediately.
Availablefrom:M.A.Hidalgo
Retrievedon:28December2015
A study of the magnetotransport properties of the graphene (II. Fractional Quantum
Hall Effect)
M. A. Hidalgo
Departamento de Física y Matemáticas
Universidad de Alcalá
Alcalá de Henares, Madrid, Spain
Correspondence and request for materials should be addressed to [email protected]
Abstract
We present an approach to the fractional quantum Hall effect observed in grapheme
(GFQHE), basing us on the model developed previously for the fractional quantum Hall
effect in a two-dimensional electron system embedded in a quantum well (FQHE) [Hidalgo,
2013(*)]. The main idea in the view proposed f0r the FQHE is the breaking of the symmetry
of the Hamiltonian of every electron in a two-dimensional electron gas (2DEG) under the
application of a magnetic field and in the presence of an electrostatic potential due to the
ionized impurities. As the magnetic field increases the effect of that electrostatic potential
evolves; changing in turn the spatial symmetry of the Hamiltonian: from continuous to
discrete one. The model provides the odd fractional states, and corresponding plateaux, p/3,
p being any integer, observed in graphene.
(*) All the references from the author can be downloaded from his “Researchgate” website
Introduction
The FQHE is known since more than thirty decades ago. After the pristine measurements by
Tsui et al., [1982], who observed the appearance of the 1/3 plateau, the importance of the
phenomenon were confirmed with the observation in any two dimensional electron system
(2DES) embedded in quantum well semiconductors (QW), of large series of other fractional
plateaux, most of them odd: {4/5, 2/3, 3/5, 4/7, 5/9, 5/3, 8/5, 11/7, 10/7, 7/5, 4/3, 9/7} [Clark
et al., 1986]; {5/3, 8/5, 10/7, 7/5, 4/3, 9/7, 4/5, 3/4, 5/7, 2/3, 3/5, 4/7, 5/9, 6/11, 7/13, 6/13,
5/11, 4/9, 3/7, 2/5, and also 8/3, 19/7, 33/13, 32/13, 7/3, 16/7}[Willet et al., 1987, 1988];
{2/3, 7/11, 3/5, 4/7, 5/9, 6/11, 5/11, 4/9, 3/7, 2/5, 1/3, 2/7, 3/11, 4/15, 3/13, 2/9} [Du et al.,
1988]; {14/5, 19/7, 8/3, 13/5, 23/9, 22/9, 17/7, 12/5, 7/3, 16/7, 11/5} [Choi et al., 2007];
{19/5, 16/5, 14/5, 8/5, 7/3, 11/5, 11/3, 18/5, 17/5, 10/3, 13/5, 12/5} [Shabani et al., 2010].
Although sometimes, but with much less frequency, even planteaux: {15/4, 7/2, 13/4, 11/4,
5/2, 9/4} [Clark et al., 1986]; {11/4, 5/2, 9/4} [Willet et al., 1987]; {1/4} [Willet et al., 1988];
{11/4, 21/8, 5/2, 19/8, 9/4} [Du et al., 1988]; {7/2, 5/2} [Choi et al., 2007].
The phenomenon shows their main features in the data related to magnetotransport in a two-
dimensional electron system (2DES): minima or zeroes in the longitudinal resistance, i.e.
Shubnikov-de Haas effect (SdH), and well-defined plateaux in the Hall resistance (at the
fractional values of the fundamental Hall resistance RH = h e2 ).
Concerning on its physical explanation, the FQHE was interpreted in the beginning as a
consequence of the formation of correlated states in the electron liquid behavior of the
2DES, [Laughlin, 1983]. In fact, Laughlin proposed that the origin of the observed FQHE
1/3, -as well any 1/q, with q being an odd integer-, is due to the formation of a correlated
incompressible electron liquid; and from that view the electron-electron interaction is
analyzed constructing several electron wave-functions ad hoc, i.e. therefore, theoretically
all the objective in analyzing the phenomenon is searching for explicit trial wave functions
corresponding to the states of the 2DES that does not break any continuous spatial symmetry
and show energy gaps. The nature of these states is related to uniform density condensates.
Because after the discovery of the 1/3 plateau, many other fractional plateaux have been
observed, following the general expression p/(2sp±1), with s and p being integers, the initial
idea by Laughlin has more recently been elaborated, assuming the appearance of new
composite quasiparticles of fractional electronic charges, coming from the combination of
electrons and flux quanta, the so-called composite fermions, according to which the FQHE
may be viewed as an integer quantum Hall effect (IQHE) of quasiparticles consisting of an
electron capturing an even number of flux quanta [Jain, 1992]. These composite entities
become elementary particles of the system.
Analyzing all the experiments related to FQHE on different 2DESs, the fundamental
character of the FQHE phenomenon is obvious, seeming clear the common origin of the
physics underneath. (In this context, it is important to highlight the fact that in all FQHE
measurements the integer plateaux are also observed.)
In fact, a completely different approach to the FQHE has recently been proposed, [Hidalgo,
2013], in which the phenomenon appears to be a physical consequence of the effect of the
interaction of electrons with impurities in the quantum limit condition.
On the other hand, in a previous work, [Hidalgo, 2014], the authors present an alternative
single electron approach to the integer quantum Hall effect in monolayer graphene
(GIQHE), natural extension of the model developed for the integer quatum Hall effect
(IQHE) in QW semiconductor, [Hidalgo, 1995, 1998, 2007]. It reproduces all the observed
characteristics of the magnetotransport properties in the monolayer graphene (MLG), and
integrating it in the same framework as the IQHE and the FQHE: the same first physical
principles, hypothesis and assumptions. The only different aspect between both two
dimensional electron systems, QW and graphene, would be the quantization of two bands,
the VB and CB, in that last case.
The fractional quantum Hall effect in graphene (GFQHE), also appear to be so fundamental
as the FQHE in QW. However, the theoretical explanation is nowadays an open question.
The present work wants to be a simple theoretical approach to the problem, extension of our
previous model for the FQHE in QW, [Hidalgo, 2013], to the GFQHE.
Origin of fractional odd denominator quantum levels in graphene
MLG is a flat layer of carbon atoms arranged in a hexagonal lattice with two carbon atoms
per unit cell. Of the four valence states, three sp2 orbitals form a σ state with three
neighbouring carbon atoms, and one p orbital develops into delocalized π and π* states that
provide the occupied valence band (VB) and the lowest unoccupied conduction band (CB).
We now consider εVB and εCB as the corresponding Hilbert spaces associated with the
dynamic states of each band, valence and conduction. We assume the same effective masses,
i.e., mVB* =mCB
* =m* , where mVB* and mCB
* would correspond to the effective masses of the
valence and conduction bands, respectively.
As we have already mentioned above, our approach to the GFQHE will be based on the
same assumptions and single electron approach as the IQHE and FQHE in QW
semiconductor, [Hidalgo 1995, 2013]. Thus, our first task will be to determine the quantized
electron states, taking into account the symmetric gauge!A= B !y,x( ) 2 , being
!B = 0,0,B( )
the magnetic field, assumed to be perpendicular to the plane defined by the graphene. In the
present case, the resultant Hilbert space for the dynamical states of any electron under the
application of the magnetic field will be the tensor product of the dynamical spaces of each
band, i.e., ! B = !VBB !!CB
B , where !VBB and !CB
B are the corresponding dynamical spaces for the
valence and conduction bands under the presence of B.
Therefore, due to the contribution of the two bands, the electrons will have two degrees of
freedom and, then, the quantized energy spectrum under the application of B has to be
described by that corresponding to an isotropic bidimensional harmonic oscillator. Hence,
the quantized levels will be given by the expression
En = i + j( )!!0 = n!!0 = nE0
(1)
with i, j, natural numbers related to each dynamical space, n=0, 1, 2…, !0 = eB m* the
fundamental angular frequency, and E0 = !!0 . These levels are degenerate in all possible
angular momentum states, determined by m. Another consequence of ! B = !VBB !!CB
B , is that
the module of the wave number has now to be
k = kCB + kVB
On the other hand, we know that the expected value of the square of the distance from the
center of the trajectory of the electron to the origin -in our case each ionized impurities,
Figure 1-, for each band, referenced in all below as Larmor radius, RO, is given by
< RO2 >= qR2 (2)
being q an odd number, and R = ! eB , the magnetic wavelength. Hence, we can define the
wavelength associated with this Larmor radius as
!q = !qCB = !q
VB = 2! < RO2 > (3)
(that is expected to be the same for both bands). Thus, we have now for the case of graphene
kq = kqCB + kq
VB = 2kqVB
(4)
that can be written as
kq = 2 2!"q
=2eBq!
(5)
It is well-known that the general Hamiltonian of every electron in the 2DEG is
H =12m*
!p+ e!A( )
2+U !r( ) = 1
2m*!p+ e!A( )
2+U i !r( )+U e !r( ) (6)
The last two terms correspond to the energy contribution of the electrostatic potentials due
to the ionized impurities, U i !r( ) , and the electron-electron interaction, U e !r( ) . (Later we
will discuss about the Zeeman term.) Of course, we consider that the electrostatic potentials
are perturbations respect to the effect of the magnetic field over the electrons.
Now, below, we summarize the model, although a detail description of it can be found in
Hidalgo [2013].
At low magnetic field the Larmor radius of every electron is large; then their interactions
U i !r( ) with the ionized impurities, because in that case we have di << 2RO , RO given by
Equation (2) and di being the distance among ionized impurities, can be assumed to be a
uniform term, and thus, the energy states for every electron areEn = E0 +U0 . (The
electrostatic potential associated with the electron-electron interaction can be neglected in
average due to the symmetry related to the electron distribution in the 2DES.)
However, at high magnetic fields the Larmor radius will be of the order of di, and, then, the
effect of the ionized impurities over every electron will now contribute to the Hamiltonian
with a non-uniform term U i !r( ) , what involves a change in the symmetry of the Hamiltonian
of the 2DEG, leaded from the continuous spatial symmetry at low magnetic field to a
discrete one at high magnetic fields. Therefore, under this last condition, we can view the
new states as an arrangement of cyclotron orbits according to the expected symmetry of the
distribution of the ionized impurities, Figure 1. To characterize this new symmetry we
introduce a correlation length related to that short-range order, which we will express as ηdi.
Then, this new symmetry implies that H r( ) = H r +!"di( ) , where α is a natural number.
Taking into account the Landau functions, {!nm }, associated with the cyclotron orbits, we
can express their spatial correlation through the relation, [Hidalgo, 2013],
<!nm r +!"di( ) / H2 / !n
m r + ! ±1!" #$"di( ) >= ±! 2 (7)
where we suppose that γ is the same for all cyclotron orbits. (We also assume the higher
correlations terms negligible.) The Landau functions does not verify the Bloch theorem, but
we can construct a new base reflecting that new short range order taking the linear
combination of the cyclotron orbits functions, [Hidalgo, 2013].
Under all these conditions, the new energy states can be expressed by the equation
E = En ! ! ±!2kq!di( )
2 (8)
from where it is easy to deduce that ! =U0 . Then, if we write En = En !U0 , we arrive to the
relation
E = En ±!2kq"di( )
2= En + Eq
" (9)
with Eq! = ±!2
U0
2kqdi( )
2. Considering the effective mass m* = !2
U0 di2 , and using (5) we can
write
Eq! = ±
!2
qE0
(10)
Hence, the energy states are
E = En ±!2
qE0 = nE0 ±
!2
qE0 +U0 = n±!
2
q
!
"#
$
%&E0 +U0 (11)
As we have described above, the expected correlations in the arrangement of the electron
cyclotron orbits in the 2DES in graphene is like that shown in Figure 1, short range order to
the first neighbors, which correspond to values of the correlation length, di, i.e., η=1.
Therefore, we have the energy states for every electron
E!!0
= n± 1q
!
"#
$
%& (12)
In Table I we detail the values of Equation (12) for the Landau levels n=0, 1 and 2, and q=3.
As it is seen, the values obtained coincide with the family of odd plateaux observed in the
experiments of the GFQHE, [Bolotin et al., 2009], [Du et al., 2009], [Gharari et al., 2011],
[Dean et al., 2011].
Finally, the contribution of the Zeeman and the spin-orbit coupling terms is summarized in
the expression Espin = g *m*!!0 4m , being g* the generalized gyromagnetic factor. In the
FQHE (and also GFQHE) conditions we can consider that all the electrons are uniformly
polarized.
To illustrate the results of the model we will show the simulation of both
magnetoconductivities (SdH and Hall effects), for any 2DEG in graphene under fractional
quantum Hall conditions. Again, the detail description of the general procedure can be found
in Hidalgo, [2013]. Here we only summarize it.
The density of states for q is then given by
gq E( ) = g0 1+ 2 A!, pcos2!qpE!!0
"g*4m*m
#
$%
&
'(
)
*+
,
-.
p=1
/
0123
43
563
73 (13)
where ⎭⎬⎫
⎩⎨⎧−= 22
2222
cp,
pexpAωΓπ
Γ , is the gaussian term related to the width of the energy
levels, and due to the interaction of electrons with defects and impurities, [Ando et al.,
1982]. (For sake of simplicity we have assumed gaussian width for each energy level,
independent of the magnetic field.)
From the density of sates the electron density is easily obtained, [Hidalgo, 1995, 1998],
n = n0 +!n = n0 +2eBhq
1! p
A!, psen Xq,F "g*
4m*m
#
$%
&
'(
p=1
)
* (14)
with Xq,F = 2!qpEF !!0 . (In this expression we have supposed very low temperature, the
ideal experimental condition to observe the FQHE.) EF is the Fermi level of the 2DEG,
given by
EF =2!!2nem*
(15)
where ne is the electron density of the 2DEG, and m* the electron effective mass.
Thus, from equations (13) and (14) the magnetoconductivities can be calculated, [Hidalgo
1995, 1998, 2007]. For testing and comparing our model we have used the experimental
measurements by Dean et al., [2011]. In Figure 2 we show the results at high magnetic field
(35 T) for the Hall magnetoconductivity, (a), and both magnetoresistivities, (b), as a function
of the gate voltage, and for the series corresponding to the odd denominator q=3.
Study of the experimental maxima of the SdH oscillations
From the experimental SdH oscillations of any GFQHE observation, one can check that the
maxima of every oscillation match with the values of the magnetic field given by the
relation, [Hidalgo, 1995, 2013, 2014],
Bq ! 3m*EF!e
(16)
EF given by (15). Even more, each fraction p/3 corresponds to the plateau with the same
fractional number.
Summary and Conclusions
We have presented a different approach to the GFQHE, being the change in the symmetry of
the Hamiltonian of any electron of the 2DEG, when the magnetic field is increasing in the
presence of the electrostatic potential due to the ionized impurities, the responsible of it. As
the magnetic field increases the effect of that electrostatic potential evolves; changing in turn
the spatial symmetry of the Hamiltonian: from continuous to discrete one. Therefore, in the
case of the graphene the series of fractions obtained {1/3, 2/3, 1, 4/3, 5/3, 2, 7/3, 8/3, 3, 10/3,
11/3, 4, 15/3}. And this states are consequence of the appearance of a correlation among the
cyclotron orbits with correlation index η=1, [Hidalgo, 2013].
Therefore the model justifies all the odd q=3 plateaux observed in the experiments on
graphene. Besides, the present work shows that its physical origin is the same than this
observed in QW semiconductors.
Thus, we hope that the shown approach could be a good starting point to understand the
magnetotransport phenomena related to graphene.
References
1.- Ando, T., Fowler, A. B., Stern, F. Electronic Properties of Two-dimensional Systems
Reviews of Modern Physics 54, 437-672 (1982).
2.- Bolotin, K., Ghahan, F., Schulman, M. D., Stormer, H. L., Kim, P. Observation of the
fractional quantum Hall effect in graphene. Nature 462, doi:10.1038/nature085582 (2009).
3.- Choi, H. C., Kang, W., Das Sarma, S., Pfeiffer, L. M., West, K. W. Fractional quantum
Hall effect in the second Landau level. Arxiv: 0707.0236v2 (2007).
4.- Clark, R. G., Nicholas, R. J., Usher, A., Foxon, C. T., Harris, J. J. Odd and even
fractionally quantized states in GaAs-GaAlAs hetrojunctions. Surface Science 170, 141-147
(1986).
5.- Dean, C. R., Young A. F., Cadden-Zimansky, P., Watanabe, K., Taniguchi, T., Kim, P.,
Hone, J., Shepard, K. L. Multicomponent fractional effect in graphene. Nature Physics 7,
693, (2011).
6.- Du, R. R., Störmer, H. L., Tsui, D. C., Pfeiffer, L. N., West, K. W. Experimental
evidence for new particles in the fractional quantum Hall effect. Phys. Rev. B 70, 2944-2947
(1988).
7.- Du, X., Skachko, I., Duerr, F., Luican, A., Andrei, E. Y. Fractional quantum Hall effect
and the insulating phase of Dirac electrons in graphene. Nature 462,
doi:10.1038/nature08522 (2009).
8.- Gharari, F., Zhao, Y., Cadden-Zimansky, P., Bolotin, K., Kim, P. Measurement of the
ν=1/3 fractional quantum Hall energy gap in suspended graphene. Phys. Rev. Lett. 106,
046801 (2011).
9.- Hidalgo, M. A. Contribución al estudio de los efectos galvanomagnéticos en el gas de
electrones bidimensional. PhD Thesis. Editorial de la Universidad Complutense de Madrid
(1995). (*)
10.- Hidalgo, M. A. A semiclassical approach to the integer quantum Hall problem.
Microelectronic Engineering 43-44, 453-458 (1998). (*)
11.- Hidalgo, M. A., Cangas, R. A model for the study of the Shubnikov-de Haas and the
integer quantum Hall effects ina a two dimensional electronic system. Arxiv: 0707.4371
(2007). (*)
12.- Hidalgo, M. A. A study of the fractional quantum Hall effect: Odd and even
denominator plateau. Arxiv: 1310.1787 (2013). (*)
13.- Hidalgo, M. A. A study of the magnetotransport properties of the graphene (I.
Monolayer). ArXiv 1404.5537 doi:10.13140/2.1.2006.5923 (2014). (*)
13.- Jain, J. K. Microscopic theory of fractional quantum Hall effect. Advances in Physics
41, 105-146 (1992).
14.- Laughlin, R. B. Anomalous quantum Hall effect: An incompressible quantum fluid with
fractionally charged excitations. Phys. Rev. Lett. 50, 1395-1398 (1983).
16.- Shabani, J., Shayegan, M. Fractional quantum Hall effect at high fillings in a two-
subband electron system. Arxiv: 1004.09/9v1 (2010).
17.- Tsui, D. C., Störmer, H. L., Gossard, A. C. Two-dimensional magnetotransport in the
extreme quantum limit. Phys. Rev. Lett. 48, 1559-1562 (1982).
18.- Willet, R. L. et al. Observations of an even-denominator quantum number in the
fractional quantum Hall effect. Phys. Rev. Lett. 59, 1776-1779 (1987).
19.- Willet, R. L. et al. Termination of the series of fractional quantum Hall states at small
filling factors. Phys. Rev. B 59, 7881-7884 (1988).
(*) All the references from the author can be downloaded from his “Researchgate” website
Acknowledgments:
The author would like to thank R. Cangas for valuable discussions
Table I: Fractional energy states for correlation length η=1.
0
E!!
n! 1q
"
#$
%
&'
n + 1q
!
"#
$
%&
n=1 n=2 n=3 n=4 n=0 n=1 n=2 n=3 n=4
q=3 23
53
83
113
13
43
73
103
133
All the fractions shown in the table have already been observed in the experiments, [Bolotin
et al., 2009], [Du et al., 2009], [Dean et al., 2011], [Gharari et al., 2011]. n corresponds to
the Landau levels. The main odd denominator fraction usually observed in graphene is
related to correlation length 1=η . (In Hidalgo [2013] can be found a detailed descripton of
the model for the FQHE in QW semiconductors, where several series of the fractional
plateaux are measured.)
Figure legends:
Figure 1: Sketch with an artistic picture of the expected short-range order in a two-
dimensional electron system due to an ionized impurities distribution underneath, and
the arrangement of the corresponding electron cyclotron orbits
Expected close packing arrangement of identical electron cyclotron orbits of a 2DEG in
fractional quantum Hall conditions. In correspondence with Table I, the correlation detailed
in the picture corresponds to a correlation length 1=η . (With black points are represented
the impurities. di is the mean distance among the ionized impurities.)
id
Figure 2: Model’s simulation of the fractional quantum Hall magnetoconductivity and
both magnetoresistivities as a function of the gate voltage. It is shown the set of odd
denominator states q=3, and the corresponding plateaux.
We have simulated with our model the fractional quantum Hall effect in the conditions of
the measurements by Dean et al. [2011]. The effective mass used in the simulation is
0.0125m, T=0.3 K, g=2, and Γ=1.8×10-20 J. a) Hall magnetoconductivity where the plateaux
have been highlighted. b) Both magnetoresistivities.